The SaaS platform labels digital advertising 200x faster than manual annotation
Deep.ad, the only computer vision company that recognizes brands, trademarks, and advertising in any type of media, announced the launch of PanelBot, a SaaS data extraction platform for market research companies.
Marketing Technology News: Award-Winning Marketing Leader Joins Uptime.Com As Chief Marketing Officer
Combining state-of-the-art object detection and optical character recognition (OCR) technology, PanelBot labels brandprints that appear within content notable to marketers, including paid media creatives and user-generated content.
As brands seek new ways to understand digital-first and social-first consumers, the number of market research companies offering first-party panel data is growing exponentially. Collecting online browsing and purchasing behavior from consumers is typically simple for these companies; millions of people are willing to exchange data for dollars. But tagging and analyzing visual media yielded from panels can be a herculean effort.
“PanelBot is the automated labeling platform market research companies have been waiting for,” said Deep.ad CEO Kristopher Kubicki. “It’s the fastest way to go from a black hole of unstructured and unanalyzed rich media to objective and actionable market intelligence.”
PanelBot’s reporting output answers two key questions: 1) What market intelligence – including brand names, logos, product categories, trademarks, pricing, promotions, etc. – does a scanned media asset contain? and 2) Where is the asset from – Facebook, YouTubeTV, TikTok, Google, Amazon, Twitter, etc. – and what else is publicly known about the content publisher?
“99% of new digital content is image or video, including paid media. That’s a scary blind spot for brands who want to know how their products are being shared in the wild and how their competitors are advertising,” said Kubicki. “PanelBot is the scalable alternative to spending a fortune on low-quality manual OCR to address this gap.”
In addition to raw image, video, and gif files and URLs, PanelBot also scrapes PDFs, audio files, photographs, HTML pages, and other media. Compared to manual annotation, it reduces labeling costs by 70% and lead time by 99%.
Marketing Technology News: NVIDIA Inference Breakthrough Makes Conversational AI Smarter, More Interactive From Cloud To Edge